Why does Net Promoter Score (NPS) matter?
NPS is popular because it's cheap to collect and easy to trend, and that's exactly what makes it dangerous to over-read. It's a stated intention, not an observed behavior. The same gap that makes leading questions unreliable in customer interviews applies here, just at scale. A rising NPS with flat or falling retention should be trusted less than the retention curve, because retention is what people actually did, not what they said they might do. NPS is worth collecting as a fast directional signal and worth distrusting as a proof point on its own.
What does Net Promoter Score (NPS) look like in practice?
Suppose a survey of 100 users returns 50 promoters, 30 passives, and 20 detractors, an NPS of 30. That number alone says almost nothing about whether the business is healthy. Segmented, it might reveal that promoters are concentrated in the practices with the highest retention and detractors in the segment already known to be churning, in which case NPS is confirming something the retention data already showed. If NPS is rising while cohort retention is flattening, the honest read is that people are getting more polite, not that the product is getting better.
What are the common mistakes with Net Promoter Score (NPS)?
- Treating NPS as a proxy for retention rather than checking whether it actually correlates with retention in this business.
- Reporting a single blended score instead of segmenting it, which hides that one segment is dragging the average down.
- Chasing NPS improvements with cosmetic changes that make people feel more positive without changing whether they stay.
- Sampling only engaged users who respond to surveys, which silently excludes the detractors most likely to already be gone.
Where the term comes from
Fred Reichheld of Bain introduced the score in a December 2003 Harvard Business Review article whose title is the whole argument: "The One Number You Need to Grow". The 0-to-10 scale and the promoter/detractor split were designed so a company could track word of mouth with a single question, and the surrounding claim, that this one number predicts growth, has been contested by researchers ever since.
Harvard Business Review, December 2003 ↗Related concepts
- Cohort RetentionThe share of a group of users who started at the same time and are still active after a given period, measured per group rather than across the whole user base.
- ChurnThe rate at which customers stop paying you, counted either as customers lost (logo churn) or as revenue lost (revenue churn), which can differ sharply.
- ActivationThe point at which a new user first experiences the product's core value, not signing up, not logging in, but doing the specific thing that makes them understand why the product exists.